Want to know if this job is worth applying to?
Fort Belvoir, Virginia, United States
Onsite
Position Summary
The AI SME provides technical and policy advisory support on artificial intelligence technologies, their applicability to arms control verification and implementation, and the potential treatment of AI within future arms control agreements, treaties, or risk-reduction measures. The position supports GT-AT and the broader Directorate on a task-directed basis and contributes to technical analysis, knowledge management, interagency engagement, and development of future arms-control concepts involving AI.
Advise DTRA on the current state and projected trajectory of AI capabilities relevant to treaty monitoring, on-site inspection, verification, imagery analysis, data fusion, anomaly detection, and compliance assessment.
Assess risks from adversary use of AI to undermine verification measures, spoof monitoring systems, or accelerate WMD and delivery-system programs.
Identify opportunities to use AI tools to improve arms-control implementation, training effectiveness, and knowledge management.
Evaluate the technical feasibility, verification challenges, and policy implications of potential arms-control frameworks addressing AI development, deployment, compute, hardware infrastructure, or application domains.
Assess proposed AI-related arms-control measures involving compute thresholds, model capabilities, dataset size, autonomous weapons, or similar technical constraints.
Analyze lessons from prior arms-control frameworks, including START, INF, and the Chemical Weapons Convention, for applicability to AI governance.
Advise on semiconductor manufacturing, export controls, supply chains, compute monitoring, and related verification mechanisms.
Support development of technical briefings and USG positions for interagency and international engagements on AI risk-reduction measures.
Maintain AI technical references, policy documents, and analysis products on Government SharePoint and restricted-drive repositories.
Prepare periodic AI landscape summaries covering significant AI capability developments, international governance efforts, and dual-use risks as directed by the COR.
Support development and maintenance of SOPs and best practices for integrating AI tools into arms-control training and implementation activities.
Attend technical demonstrations, interagency workshops, academic/industry conferences, and other relevant engagements as directed.
An in-depth understanding of large language models (LLMs) and machine-learning architectures, training methodologies, fine-tuning, evaluation, and deployment at scale.
Knowledge of AI hardware and software infrastructure, including GPU/TPU ecosystems, data-center architecture, semiconductor supply chains, distributed training, and inference optimization.
Familiarity with major AI development stacks and frameworks such as PyTorch, TensorFlow, and JAX, as well as model evaluation, benchmarking, AI safety, and alignment research.
Current understanding of frontier-model capabilities, multimodal systems, autonomous agents, dual-use AI risks, and the global competitive AI landscape.
Experience with AI governance, policy, or international regulatory frameworks and engagement with government, academic, or industry stakeholders.
Prior national security, defense, arms-control, verification, WMD, or strategic-risk-reduction experience.
Experience translating highly technical AI concepts into concise analysis and decision support for senior Government leaders.
Experience working across interagency, academic, and industry communities.
Technical analyses, advisory products, briefings, and recommendations assigned by the COR.
AI landscape summaries and technical reference repositories.
Knowledge-management products, SOPs, and best-practice documentation.
Support to interagency and international engagement materials.
